• A Lightweight Object Detection Network called LUD-YOLO for UAVs.
• A new feature fusion pattern has been proposed to address the degradation of feature interaction.
• Proposing a novel feature extraction module to improve inference ...

Autonomous execution of tasks by unmanned aerial vehicles (UAVs) relies heavily on object detection. However, object detection in most images presents challenges such as complex backgrounds, small targets, and obstructions. Additionally, the ...

Highly detailed and accurate segmentation and classification of images constitutes an important class of tasks in computer vision. Typical “universal” domain-agnostic methods are known to suffer from instabilities and are prone to adversarial ...

• Proposed methodology can be applied on real-life biomedical data.
• Developed the pipeline for automated feature selection from convoluted complex histopathological dataset.
• Several algorithms were tested on histopathological slides ...

We consider a long-standing yet hard and largely open machine learning problem of anomaly areas detection in multimodal 3D images. Purely data-driven methods often fail in such tasks because rarely incorporating domain-specific knowledge into the ...

Video semantic segmentation (VSS) is a fundamental machine vision task with various practical applications, such as autonomous driving and automated surveillance. Current studies mainly utilize temporal features based on optical flow and the self-...

• The model TDSNet is constructed to accurately segment frames of the video.
• The temporal-difference-based temporal features are utilized to improve accuracy of the segmentation.
• The temporal features are adaptively aggregated ...

Training convolutional neural networks (CNNs) using back-propagation (BP) is a time-consuming and resource-intensive process, primarily due to the need to iterate over the dataset multiple times. In contrast, analytic learning aims to train ...

• A fresh analytic formulation (ACnnL) for training CNNs unlike the iterative strategies.
• The CNN structure can be treated as a generalized MLP with significantly more constraints.
• The ACnnL completes the training with a ...

The objectives of infrared and visible image fusion are to generate a single image that includes significant objects and rich texture information. However, the current deep-learning methods ignore uncertainty in the decision-making process during ...

• Proposed a novel infrared and visible image fusion model.
• The Swin Transformer module used for feature extraction.
• Fusion process established with intuitionistic fuzzy sets and knowledge measures.
• Develop general construction ...

Images generated by the Generative Adversarial Network (GAN) are too realistic to be distinguished by humans. Recently, some detection methods have been proposed to distinguish between generated and real images. However, these methods rely on ...

• A disentangled representation framework for evading GAN-generated image detection.
• To conceal a fake image, replace its fake features with those of real images.
• The method's training is detection-independent and applicable to many ...

Given the significant differences between domains, current unpaired learning methods struggle to accurately map the relationship between rainy and clear images. To this end, we introduce a neural Schrödinger bridge (NSB) for unpaired real-world ...

Tensor data, such as hyperspectral images and multi-frame videos, have gained significant attention in practical applications. However, the inherent degradation phenomena during data acquisition, including noise and missing pixels, give rise to a ...

Image segmentation allows us to separate an image into distinct, non-overlapping parts by utilizing specific features such as hue, texture, and shape. The technique is prevalent in different domains, including target detection, medical imaging, ...

• We use an IFS to represent uncertainty, combine local information propose a novel image representation method.
• We propose an exponential distance measure for IFSs, and show its superiority by illustrating examples.
• We develop a ...

Person search is an extremely challenging task that seeks to identify individuals through joint person detection and person re-identification from uncropped real scene images. Previous studies primarily focus on learning rich features to enhance ...

Neighborhood rough sets-based methods have been widely used for feature selection. However, the existing methods have some problems in neighborhood construction, such as the application of the same neighborhood radius for all samples. Thus, this ...

In this paper, our objective is to simultaneously explore the learning of ordinal relationships among age labels and address the challenge of heterogeneous data resulting from the non-stationary aging process through an advanced mixture model of ...

• We propose a new deep mixture model that learns ordinal age relationships and handles the non-stationary aging process simultaneously.
• Two well-designed visual tasks are proposed for ordinal relation learning and age estimation.
• A ...

For the purpose of monitoring fish health, managing aquaculture, and comprehending marine ecology, there is a growing interest in the automatic classification of different fish species. Recent developments in machine vision-based classification ...

This paper delves into the realm of object detection models, pinpointing challenges posed by inadequate performance in discerning small objects, as well as examining the inherent imbalance between positive and negative samples. To address these ...

Domain generalization (DG) aims to train a model capable of generalizing to unseen target domains by utilizing data from multiple disjoint source domains. Previous domain generalization methods primarily utilize data augmentation and domain-...

The health status of the retinal blood vessels is a significant reference for rapid and non-invasive diagnosis of various ophthalmological, diabetic, and cardio-cerebrovascular diseases. However, retinal vessels are characterized by ambiguous ...

Subspace projection has been widely studied to implement the feature extraction for face recognition. However, it scarcely facilitates exploring the label information in the subspace projection procedure. The feature extraction and classification ...

Although deep learning models can solve complex prediction problems, they have been criticized for being ‘black boxes’. This implies that their decisions are difficult, if not impossible, to explain by simply inspecting their internal knowledge ...

• Cascaded multiresolution design is introduced into 3D convolution.
• The algorithm needs to have strong anti-interference ability to light changes.
• Micro-expression changes can be captured by deep attention mechanism.

Virtual reality/augmented reality (VR/AR) technology is very important for meta-cosmic medicine. It is the technical link between the material world and the virtual world. The realization of face-to-face psychological diagnosis with three-...